Accuracy of Ultrasonography in the Diagnosis of Peritonitis Compared with the Clinical Impression of the Surgeon

ORIGINAL ARTICLE Accuracy of Ultrasonography in the Diagnosis of Peritonitis Compared With the Clinical Impression of the Surgeon

Shyr-Chyr Chen, MD; Fang-Yue Lin, MD, PhD; Yeu-Sheng Hsieh, PhD; Wei-Jao Chen, MD, PhD

Hypothesis: Peritonitis is a well-known indication for Results: Ultrasonography and clinical impression surgery, but its preoperative cause usually is not estab- accurately diagnosed the peritonitis in 85 (83.3%) and lished. We hypothesize that abdominal ultrasonogra- 52 (51.0%) of the patients, respectively. The difference phy is superior to the clinical impression of the surgeon between ultrasonography and clinical impression in the in detecting the cause of peritonitis. diagnosis of peritonitis was significant (PϽ.001). Among 45 patients without a preoperative clinical diag- Design: A prospective case series. nosis, a diagnosis was made by ultrasonography for 32 (71%) of them. There were a total of 98 patients with Setting: A major university hospital in Taiwan, Repub- positive ultrasonographic findings, and 4 patients had lic of China. normal screening results. Of the 98 patients with positive ultrasonographic findings undergoing surgery, all Patients and Methods: One hundred two patients had abdominal pathological characteristics. The 4 with a diagnosis of peritonitis admitted to the Depart- patients with normal screening results received nonop- ment of Emergency Medicine, National Taiwan Univer- erative treatment. sity Hospital, Taipei, were included in this study. All 102 patients underwent an abdominal ultrasonographic Conclusions: Ultrasonography is a more sensitive tech- examination; and the ultrasonographic findings of these nique than clinical judgment in diagnosing peritonitis. patients were classified into 2 categories: positive find- Ultrasonography may be a useful diagnosing modality ings and normal screening results. The accuracy of in patients with peritonitis in whom the clinical cause is clinical impression in detecting the cause of peritonitis unclear. was compared with the accuracy of abdominal ultrasonography. Arch Surg. 2000;135:170-173

CONCERNING fraction of morbidity and prolonged hospitaliza- patients who present with tion.1-5 To reduce the rate of unneces- acute abdominal pain have sary laparotomy, additional sensitive peritonitis. Peritonitis re- and specific examinations are needed to fers to any inflammation of Athe peritoneal layers, and it is an emer- See Invited Critique gency condition that frequently requires surgery. Not every underlying cause of at end of article peritonitis is diagnosed before surgery, and laparotomy has, therefore, traditionally screen patients for operative indications been advised to treat the patient’s disease before surgery. Plain radiography, lapa- and to determine the nature of the ab- rotomy, computed tomography, and dominal pathological features. However, ultrasonography have been used to aid From the Departments of some medical conditions may mimic acute in the diagnosis of peritonitis. To our Emergency Medicine peritonitis, which do not require surgical knowledge, the role of the abdominal (Drs S.-C. Chen and Lin) and intervention. Should surgeons operate im- ultrasonographic examination in patients Surgery (Dr W.-J. Chen), mediately and face the possibility of find- with peritonitis has not been well re- National Taiwan University Hospital, and the Department of ing a nonsurgical condition at lapa- ported. This prospective study compared Agricultural Extension, rotomy, or should they observe and risk the diagnostic accuracy of abdominal National Taiwan University missing the optimal time for surgery? ultrasonography with the clinical im- (Dr Hsieh), Taipei, Taiwan, Unnecessary laparotomy has been pression of the surgeon in diagnosing Republic of China. reported to be associated with increased peritonitis.

ARCH SURG/ VOL 135, FEB 2000 WWW.ARCHSURG.COM 170

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 PATIENTS AND METHODS first screened to check the antrum, the first portion of the duodenum, and the pancreas; then screening was shifted to the right hypochondriac region and subcostal area to check All patients with peritonitis admitted to the Department of the liver, the gallbladder, and the right pleural space. The Emergency Medicine, National Taiwan University Hospi- presence of free air was checked in both regions. Follow- tal, Taipei, Taiwan, Republic of China, from August 1996 ing this, the right paracolic gutter was examined to check through March 1999, were included in this prospective the ascending colon, the terminal ileum, and the appen- study. Those patients with abdominal injury were ex- dix. Then the rectouterine pouch and the left subcostal cluded. The study included 102 patients (68 male and 34 area were investigated to check the pelvic condition, the female patients). The age of these patients ranged from 13 spleen, and the left pleural space. The left paracolic gutter to 83 years (mean, 47 years). was screened next to check the descending colon. Finally, When patients arrived at the hospital, they were ex- the central abdomen was examined to check the small amined by a member of the surgical house staff. After ob- intestine. taining a detailed medical history and performing a physi- The results of the ultrasonographic examination were cal examination, venous blood from each patient was categorized as either a positive ultrasonographic finding or sampled and a plain abdominal radiograph or a chest ra- a normal screening result. A positive ultrasonographic find- diograph was obtained. The clinical diagnosis of peritoni- ing was defined as any additional or abnormal ultrasono- tis is defined as a patient having diffuse abdominal tender- graphic changes in the abdominal cavity. The ultrasono- ness, rebounding pain, and leukocytosis. If peritonitis was graphic diagnosis was also recorded. The decision to operate diagnosed, the presumed cause was recorded, which was was based on the presence of or highly suspected surgical blinded from the ultrasonographer, and the patient then disease found by ultrasonography. Those with normal underwent an abdominal ultrasonographic examination per- screening results were hospitalized for close observation formed by a staff surgeon. This surgeon had 6 years of ex- and continued investigation. Additional tests, including pan- perience in performing abdominal ultrasonographic ex- endoscopy or computed tomography, were performed to aminations. Ultrasonography was performed with a search for an underlying cause of peritonitis in those pa- handheld 3.75-MHz curved-array transducer (model SSA- tients with normal ultrasonographic screening results. The 340A; Toshiba, Tochigi-Ken, Japan) over the whole abdo- relation between operative findings and ultrasonographic men, with screening of the pleural space, hepatorenal re- findings was examined in the assessment of the diagnostic cess, paracolic gutter, rectouterine pouch, liver, biliary tract, accuracy of ultrasonography. The statistical difference in gallbladder, spleen, pancreas, small intestine, colon, and this study was determined by the ␹2 test. PϽ.05 was con- intra-abdominal fluid collections. The epigastric area was sidered significant.

RESULTS found to have a peptic ulcer without perforation or obstruction. They received a hydrogen blocker and ant- The findings of the abdominal ultrasonographic exami- acid treatment and were followed up at the outpatient nation are shown in Table 1. The ultrasonographic di- clinic. One of these patients underwent abdominal com- agnoses of 102 patients with peritonitis are shown in puted tomography, which showed no abnormalities. She Figure 1. There were 98 patients with positive ultraso- received conservative treatment and was discharged from nographic findings and 4 patients with normal screen- the hospital 2 days later, after her abdominal pain ceased. ing results. The cause of the peritonitis in 85 patients was The symptoms of all 4 patients with normal screening accurately diagnosed by ultrasonography. Incorrect ul- results resolved without surgery, and recovery was veri- trasonographic diagnoses were found in 5 patients, in- fied at the outpatient clinic 2 weeks later. cluding 2 with pneumoperitoneum, 2 with internal bleed- Among the 57 patients with a preoperative clini- ing, and 1 with perforated appendicitis. The intraoperative cal impression (Figure 2), 5 (9%) of these diagnoses findings in patients who underwent laparotomy were as were changed based on ultrasonographic findings, follows: including 3 cases of pneumoperitoneum and 2 cases of No. of Patients perforated appendicitis. Ultrasonographic diagnoses Type of Disease were also made in 32 of the remaining 45 patients Perforated duodenal ulcer 32 Table 2 Perforated gastric ulcer 25 without a previous clinical impression ( ). In Perforated appendicitis 13 the 102 patients, an accurate diagnosis was made Colonic perforation 9 based entirely on the clinical impression in 52 patients Intestinal perforation 3 (51.0%) and based entirely on ultrasonographic Diverticulitis rupture 5 examination results in 85 patients (83.3%) (PϽ.001). Intra-abdominal abscess 5 These results clearly show that ultrasonography is a Acute cholecystitis 3 more sensitive modality than clinical judgment in Ischemic bowel 2 detecting the cause of peritonitis. Intestinal leiomyosarcoma 1 Total 98 The most common ultrasonographic findings were ascites, dilated small-bowel loop with wall thickness, The 4 patients with normal screening results re- pneumoperitoneum, thickness of the antrum or duode- ceived conservative treatment, and 3 of them under- nal wall, perforated appendicitis with perifocal exudate went an additional panendoscopic examination and were accumulation, and abscess formation.

ARCH SURG/ VOL 135, FEB 2000 WWW.ARCHSURG.COM 171

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 Table 1. Findings of an Abdominal Ultrasonographic 102 Patients Examination in 102 Patients 102 (100%) 0 With Positive With Negative No. of Clinical Findings Clinical Findings Ultrasonographic Findings Patients

Abnormal 98 57 (55.9%) 45 (41.1%) Ascites 77 With a Without a Pneumoperitoneum 62 Clinical Diagnosis Clinical Diagnosis Small-bowel edema 54 Thickness of the antrum or duodenal wall 18 52 (51.0%) 5 (4.9%) Perforated appendicitis 12 With a With an Diverticulitis rupture 4 Correct Diagnosis Incorrect Diagnosis Intra-abdominal abscess 4 Acute cholecystitis 3 Figure 2. Clinical impression of 102 patients with peritonitis. Internal bleeding 2 Pelvic mass 1 Normal 4 Table 2. A Comparison of Diagnosis Based Entirely on Ultrasonographic or Clinical Findings in 102 Patients*

Disease Ultrasonographic Clinical 102 Patients (No. of Patients) Diagnosis Diagnosis Hollow organ perforation (69) 62† 49‡ 98 (96.1%) 4 (3.9%) Perforated appendicitis (13) 12§ 8† With Positive With Negative Intra-abdominal abscess (5) 4 0 Ultrasonographic Findings Ultrasonographic Findings Diverticulitis rupture (5) 4§ 0 Acute cholecystitis (3) 3 0 90 (88.2%) 8 (7.8%) Peptic ulcer (3) 0 0 With an Without an Ischemic bowel (2) 2 0 Ultrasonographic Diagnosis Ultrasonographic Diagnosis Leiomyosarcoma rupture (1) 1࿣ 0 Internal bleeding (0) 2† 0 85 (83.3%) 5 (4.9%) Unknown (1) 12 45 With a With an Correct Diagnosis Incorrect Diagnosis Total (102) 102 102

Figure 1. Ultrasonographic diagnoses of 102 patients with peritonitis. *Data are given as number of patients. †Included 2 incorrect diagnoses. ‡Included 3 incorrect diagnoses. §Included 1 incorrect diagnosis. COMMENT ࿣Pelvic mass. Peritonitis continues to be one of the major abdominal emergencies confronting surgeons. The prognosis of peri- sive, cost-effective, easily accessible, and accurate in the tonitis is poor, especially when multiple organ failure or diagnosis of peritonitis. sepsis develops.6-9 The wide range of causes and varied Commonly performed additional imaging studies in- patient presentations pose a formidable diagnostic and clude plain radiography, laparoscopy, ultrasonography, and therapeutic challenge to the surgeon. Common causes computed tomography. These imaging modalities should include intra-abdominal inflammation, hollow organ per- be used to investigate a specific concern and not as a re- foration, trauma, bowel ischemia, and bowel obstruc- placement for clinical judgment. Plain radiographs of the tion. Regardless of the causes of peritonitis, immediate abdomen may be helpful in the examination of the patient laparotomy to achieve source control and peritoneal toil with acute abdominal pain. A single radiographic view of is the most important part of treatment for preventing the abdomen is rarely of help. The usual collection of ra- subsequent serious intra-abdominal sepsis. However, con- diographic views of the abdomen obtained includes an up- servative treatment is reserved for those patients with peri- right view; a view of the kidneys, ureters, and bladder; and tonitis that is caused by other medical conditions. Iden- an upright chest radiograph. An inspection should be per- tification of those cases of peritonitis that will not benefit formed for detecting the presence of free intraperitoneal from surgery is necessary. air, air in the retroperitoneum, patterns of gas distention, The diagnostic workup of the peritonitis always be- air-fluid levels, foreign bodies, fecaliths, and stones.10 Free gins with a precise history taking, a complete physical intraperitoneal or retroperitoneal air is a definitive diag- examination, and laboratory studies. The approach to ex- nosis of hollow organ perforation or abscess formation, and amining patients with peritonitis via diagnostic imaging the other findings are used for adjuvant diagnosis. may change when improved outcomes have been dem- Laparoscopy has been advocated as a diagnostic mo- onstrated. Inherent in the evaluation of new technology dality in the examination of a patient with acute abdomi- applied to peritonitis is how tests can avoid a misdiag- nal pain.11-13 Although directly visualizing the abdomi- nosis or provide an earlier disposition of peritonitis. The nal cavity might be a valid method of determining the goal of an imaging study as a diagnostic aid in peritoni- focus of the inflammation, patients preceded to surgery tis should be to search for the presence of surgical dis- or without any abnormal findings would be exposed to ease. In addition, it should be quick, relatively noninva- the risks and costs of general anesthesia and diagnostic

ARCH SURG/ VOL 135, FEB 2000 WWW.ARCHSURG.COM 172

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 laparoscopy. Several researchers14-17 have advocated the pathological process in cases of peritonitis but also for its use of computed tomography as the imaging test of choice ability to eliminate certain diagnoses, expedite patient care, in diagnosing peritonitis. However, computed tomogra- and improve resource allocation. In many cases, the pa- phy is associated with greater cost, exposure to ionizing tient’s medical history, clinical examination results, labo- radiotherapy, and exposure to contrast medium.18 Com- ratory findings, and the results of plain radiography of the puted tomography may more or less identify septated peri- abdomen or chest suffice to indicate the need for urgent toneal fluid, parietal peritoneum thickening, and infil- surgery, and an ultrasonographic examination is not indi- tration of the mesenteric or omental fat, fluid-filled dilated cated in these patients. However, in other cases, when clini- loop; it may also demonstrate bowel wall edema and ob- cal diagnosis is uncertain and treatment is not clear, ultra- viate the need for laparotomy.14 sonography may identify the precise cause underlying acute Abdominal ultrasonography has become a diagnos- abdominal pain that requires surgery. The results of this tic procedure of increasing importance in patients with acute study suggest that ultrasonography is a useful adjunct in a abdomen.19-24 Various reports19-21,24 have shown its benefit subgroup of patients with peritonitis whose clinical im- in surgical emergencies when a diagnosis has to be deter- pression is unclear. mined rapidly. This study was performed to evaluate the diagnostic sensitivity of abdominal ultrasonography in the Reprints: Wei-Jao Chen, MD, PhD, Department of Sur- diagnosis of peritonitis. In a previous study,25 ultrasono- gery, National Taiwan University Hospital, No. 7, Chung- graphic screening led to an earlier diagnosis, but the length Shan South Road, Taipei, Taiwan, Republic of China. of hospitalization was not reduced. In this study of 102 patients, a preoperative ultrasonographic examination accurately detected the cause of peritonitis in 85 (83.3%), REFERENCES whereas an accurate diagnosis was made for only 52 (51.0%) based on history taking, physical examination results, and 1. Renz BM, Feliciano DV. The length of hospital stay after an unnecessary laparotomy for trauma: a prospective study. J Trauma. 1996;40:187-196. laboratory findings. Thus, an ultrasonographic examina- 2. Leppaniemi A, Salo J, Haapiainen R. Complications of negative laparotomy for tion significantly increased preoperative diagnostic sensi- truncal stab wounds. J Trauma. 1995;38:54-58. 3. Renz BM, Feliciano DV. Unnecessary laparotomies for trauma: a prospective study tivity (PϽ.001). Although a preoperative diagnosis can be of morbidity. J Trauma. 1995;38:350-356. determined in some patients, ultrasonography still fails to 4. Demetriades D, Vandenbossche P, Ritz M, et al. Non-therapeutic operations for reveal the underlying cause in some patients. For ex- penetrating trauma: early morbidity and mortality. Br J Surg. 1993;80:860-861. 5. Henderson VJ, Organ CH Jr, Smith RS. Negative trauma celiotomy. Am Surg. ample, ultrasonography can demonstrate pneumoperito- 1993;59:365-370. neum, but identification of the site of hollow organ perfo- 6. Bosscha K, Reijnders K, Hulstaert PF, Algra A, van der Werken C. Prognostic scor- ration by ultrasonography is difficult. All 98 patients with ing systems to predict outcome in peritonitis and intra-abdominal sepsis. Br J Surg. 1997;84:1532-1534. positive ultrasonographic findings underwent lapa- 7. Schein M. Planned reoperations and open management in critical intra- rotomy and were found to have intra-abdominal patho- abdominal infections: prospective experience in 52 cases. World J Surg. 1991; 15:537-545. logical features that confirmed the positive ultrasono- 8. Farthmann EH, Schoffel U. Principles and limitations of operative management graphic findings. This demonstrated that ultrasonography of intra-abdominal infections. World J Surg. 1990;14:210-217. is a reliable screening modality for peritonitis and that a 9. Schein M, Saadia R, Freinkel Z, Decker GAG. Aggressive treatment of severe diffuse peritonitis: a prospective study. Br J Surg. 1988;75:173-176. normal ultrasonographic screening result excludes the like- 10. Martin RF, Rossi RL. The acute abdomen: an overview and algorithms. Surg Clin lihood of intra-abdominal pathological features. North Am. 1997;77:1227-1243. Abdominal ultrasonography in patients with perito- 11. O’Sullivan GC, Murphy D, O’Brien MG, Ireland A. Laparoscopic management of generalized peritonitis due to perforated colonic diverticula. Am J Surg. 1996; nitis provides the physician with another stethoscope to 171:432-434. check the abdominal condition. It is the single most effec- 12. Chiu CK, Karmakar MG, Yang HK, Da Silva MC, Karanfilian RG. Laparoscopic management of peritonitis in the setting of an infected Tenckhoff catheter: a case tive test for noninvasive exploration of the abdomen. The report and description of technique. Am Surg. 1996;183:640-642. reasons for performing abdominal ultrasonography in pa- 13. Eustace J, Lappin L, Keane F, Mellotte GJ, Keogh JAB. Laparoscopic-aided di- tients with peritonitis include: (1) to identify the abdomi- agnosis of recurrent peritonitis in a patient on CAPD. Nephrol Dial Transplant. 1996;11:2338-2339. nal pathological features, such as perforating appendici- 14. Taourel P, Pradel J, Fabre J-M, Cover S, Senetene E, Bruel J-M. Role of CT in the tis, acute cholecystitis, ascites, or pneumoperitoneum; (2) to acute nontraumatic abdomen. Semin Ultrasound CT MR. 1995;16:151-164. prevent unnecessary treatment or examination, such as un- 15. Taourel P, Baron MP, Pradel J, et al. Acute abdomen of unknown origin: impact of CT on diagnosis and management. Gastrointest Radiol. 1992;17:287-291. necessary laparotomy or enema study to determine the pres- 16. Stewart ET. CT diagnosis of small-bowel obstruction. AJR Am J Roentgenol. 1992; ence of toxic megacolon; (3) to provide sufficient infor- 158:771-772. 17. McDowell RK, Dawson SL. Evaluation of the abdomen in sepsis of unknown ori- mation to allow the initiation of the treatment, such as gin. Radiol Clin North Am. 1996;34:177-190. surgery or close observation and further examination; and 18. Wilcox RT, Traverso LW. Have the evaluation and treatment of acute appendi- (4) to evaluate the process of disease to determine whether citis changed with new technology? Surg Clin North Am. 1997;77:1355-1370. 19. Hudson PA, Promes SB. Abdominal ultrasound. Emerg Med Clin North Am. 1997; the disease has progressed or improved. Preoperative knowl- 15:825-848. edge of the causes of peritonitis has the advantages of en- 20. Puylaert JBCM, van der Zant FM, Rijke AM. Sonography and the acute abdo- abling the formation of a detailed operative plan, includ- men: practical considerations. Am J Radiol. 1997;168:179-186. 21. Romano WM, Platt JF. Ultrasound of the abdomen.Crit Care Clin. 1994;10:297-319. ing incision line, operative procedures, surgical instrument 22. Chen SC, Chen KM, Wang SM, Chang KJ. Abdominal sonography screening of preparation, and preoperative instruction to patients and clinically diagnosed or suspected appendicitis before surgery. World J Surg. 1998; 22:449-452. family members. With the aid of ultrasonography, physi- 23. Pattison P, Jeffrey RB, Mindelzum RE, Sommer FG. Sonography of intraabdomi- cians can easily determine the condition of the patient with nal gas collections. Am J Radiol. 1997;169:1559-1564. peritonitis and define appropriate treatment. 24. Ogata M, Mateer JR, Condon RE. Prospective evaluation of abdominal sonography for the diagnosis of bowel obstruction. Ann Surg. 1996;223:237-241. In conclusion, real-time ultrasonography is valuable 25. Gupta H, Dupuy DE. Advances in imaging of the acute abdomen. Surg Clin North not only for its sensitivity in identifying the underlying Am. 1997;77:1245-1263.

ARCH SURG/ VOL 135, FEB 2000 WWW.ARCHSURG.COM 173